Ph.D. Proposal

Christian Coletti

(Advisor: Prof. Dimitri Mavris)

"Intelligent Goal-Oriented Model-Predictive Path Integral Control (IGO-MPPI): An Autonomy Framework for UAVs and Mobile Robotics in Complex Environments"

Wednesday, July 17

9:00 a.m.

Collaborative Visualization Environment (CoVE)
Weber Space and Technology Building (SST II)
and

Microsoft Teams

Abstract

Various entities such as the Department of Defense, the Defense Advanced Research Projects Agency, the Air Force Research Laboratory, and MITRE have released research or roadmaps that demonstrate that future autonomous systems, especially UAVs, will be asked to perform increasingly complicated tasks in complex and dynamic environments. In order to improve decision-making in autonomous architectures, an emerging trajectory planning is selected that uses a forward projective, sampling-based approach: model-predictive path integral control (MPPI). In order to extend MPPI in order to handle problems requiring long-term intelligent goal-orientation (IGO), advanced simulation and cost shaping techniques are included in order to form a systematic framework used in autonomous system development. Additionally, an architecture-compatible multi-agent teaming approach is established and reviewed for viability in practical settings such as connection intermittency. This novel framework is formalized as Intelligent Goal-Oriented Model Predictive Path Integral planning (IGO-MPPI) and seeks to improve the performance of autonomous unmanned systems such as UAVs in practical problems of high complexity.

Committee

• Prof. Dimitri Mavris – School of Aerospace Engineering (advisor)
• Prof. Graeme Kennedy – School of Aerospace Engineering
• Prof. John Valasek – Aerospace Engineering Department, Texas A&M University
• Prof. Kyriakos Vamvoudakis – School of Aerospace Engineering
• Dr. Olivia Fischer – School of Aerospace Engineering
• Dr. Sean Wilson – Georgia Tech Research Institute